News from HPE NonStop Division

NonStop is not just an OS, it is a reference architecture, and a heck of a good one!

Already one month into 2023, and we’re at full speed creating new amazing NonStop products, features, and services for our existing and new customers. As highlighted at the last NonStop TBC event, we know the focus continues to be set on the cloud in all its shapes and forms but without ever losing track of our core values of availability, scalability, integrity, and security. Adding the public cloud as a new target for the NonStop platform, our journey continues “Always on, always adapting.” And with this new direction, it becomes, as usual natural to ask: What’s in it for us?

The cloud comes with a long list of paradigm shifts, new technology, and new consumption models that it feels overwhelming at times and surely intimidating. Some aspects are easy to understand, such as the service provider managing your system for you or the use of virtualization as a mechanism to consolidate your infrastructure.  On the other hand, it takes a serious effort to understand all the ramifications, which at the end of the day, will determine the quality of service, the cost, and the security level of your application.

Take, for example, the operating system. The OS used to be the center of the competition between vendors with at least a few dozen of participants. OSes would compete against each other on their feature set, their graphical interface, or their hardware support. Today, the OS is a binary question in a drop-down menu box where you select Windows or Linux.  Instead, it seems the most important question is to pick between KVM, Hyper-V, or Nitro, the so-called hypervisors that entered the stage with virtualization technology. Does this mean the OS has no differentiating value anymore? Maybe that is true between Linux and Windows but is this true for NonStop? Would you select NonStop in that same drop-down box?

As the industry is shifting to everything as a service where the OS is often managed by the service provider, at least in PaaS and SaaS, the core OS value appears to be relegated in the background even more. What matters is what the application needs. Applications typically need much more than an OS. They need an ecosystem. Maybe this is where Java and Python took some market share away from what used to be in the OS territory. The fact that Java and Python have repositories of 10’s of thousands of packages is one example of that. But I digress, the point is that “OS” is not a topic of discussion anymore in IT departments. The value is expressed in a different layer of the software stack.

For NonStop this becomes a matter of adapting. Our wonderful NSK operating system has amazing core features, process pairs, single system image at the kernel level, fail fast, virtualized resources, shared nothing, and more. So how do we highlight those attributes in a conversation where the OS has no place anymore? Part of the answer comes when you start comparing with other platforms and how they achieve the same goal yet in a very different way. You realize very often that those are not implemented at the OS level but rather by a given deployment architecture made of many assembled parts.  A load balancer is implementing an IP single system image, that’s all it does. Oracle RAC implements the clustering of Oracle databases and that’s all it does. A failover software implements a heartbeat and restart mechanism of another node, that’s all it does.  Those are not integrated together, have different scopes and require skills for each piece to assemble, or if in the public cloud, they may each be charged separately.

When we talk about NonStop, maybe we should talk about it as a reference architecture. A reference architecture that ensures high availability and scalability are built-in, that all single points of failures are removed, and that any failure does not generate any data corruption or split brains. A reference architecture that has been tested again and again in the most mission-critical environments. And as long as we can deploy that architecture into the cloud – And we saw a successful proof of concept of this done in Azure and Google Cloud at our latest TBC – then our differentiating value continues to exist just as the same even in the public cloud.

In this edition of The Connection magazine, you will find experiences of a journey to the cloud. Maybe it is not so scary after all!

Author

  • Roland Lemoine

    Roland Lemoine has been working on NonStop for 23 years and is currently the product manager for database and blockchain languages and development products. Previous experience includes customer support for middleware products, Open Source advocacy and a strong UNIX background.

Be the first to comment

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.